What is new
The reading room contains eight newly written long-form guides about sources, feeds, timestamps, market-data presentation and protocol reporting. Each carries its actual publication date and links to primary documentation. Invented examples are identified as such. No historical staff, photography, active trading service or contemporary newsroom operation has been invented to decorate the edition.
What was recovered
The archive inventory contains 8,365 HTML capture rows. This edition saved 8,359 raw responses and publishes factual metadata for 7,122 screened captures: 5,287 news records and 1,835 collection records. A capture row is not necessarily a unique article, and repeated versions are not independent reports. The public catalogue's capture range is 2021-12-10 09:38:08 UTC to 2026-02-09 03:20:11 UTC.
Raw response bytes are preserved in the local recovery corpus with SHA-256 digests. Public records retain titles, original addresses, capture identifiers, printed attribution labels where present and unambiguous publisher-article addresses where extracted. Publisher addresses are inert historical metadata. The full imported third-party article bodies and photographs are not republished.
What stays out
1,237 retrieved records were withheld by conservative screening for promotional or speculative headlines, explicit sponsorship, unrelated or insufficient article regions, or obsolete service/affiliate flows. 6 capture retrievals were unsuccessful at the recorded recovery checkpoint. Those gaps are not filled with invented text. The audit retains concrete reasons locally. Automated screening is a bounded editorial aid, not a claim that all archived reporting has been fact-checked.
Old accounts, affiliate placements, current-price simulations, offers, service contacts and biographies are not revived. Late unrelated spam does not become part of the new publication. The existing abstract artwork is used as illustration, not as historical photography, financial charts or evidence of actual market conditions.
Dates are separate fields
An archive capture identifies a saved representation. A printed byline/date line is preserved literally, without certifying its accuracy. A URL date segment is an address component, not sufficient proof of publication. New essays and new catalogue commentary use 17 October 2026; this date is not backdated to the age of the source. Historical catalogue pages do not emit Article or NewsArticle publication claims and are marked noindex.
Attribution and limits
A historical aggregator can reproduce or summarise another publisher. Its captured title is not proof of independent reporting, and a listed author is not a member of the current site's staff. Missing publisher links, authors or publication dates are left missing. No conclusion about a token, issuer, market, law or service should be treated as current solely because it appeared in an old snapshot.
Explore the evidence
Search the historical catalogue for capture-specific provenance, or read the new essays for worked explanations. The catalogue is served from a frozen local dataset; it does not depend on a runtime proxy to the former origin.